Professional Documents
Culture Documents
Obstructed Shear Ows: Similarities Across Systems and Scales
Obstructed Shear Ows: Similarities Across Systems and Scales
c Cambridge University Press 2009 51
doi:10.1017/S0022112009992175
In this paper, I show that a range of environmental flows are inherently dynamically
similar. These flows, which are partially obstructed by a permeable medium, are here
termed ‘obstructed shear flows’. Examples include aquatic flows over sediment beds,
submerged vegetation canopies and coral reefs, as well as atmospheric flows over crop
canopies, forests and cities (‘urban canopies’). While the density and geometry of the
obstructions may vary, the drag in each system generates a velocity profile with an
inflection point. This renders the flow unstable. Consequently, it is expected that (a) the
dominant interfacial turbulent structure in obstructed shear flows will be a Kelvin–
Helmholtz-type vortex, and (b) that this instability will engender hydrodynamic
similarities among obstructed shear flows. Such similarities have been hypothesized
but not yet fully explored. An extensive review of existing data confirms these dynamic
similarities on scales of O(mm) to O(10 m). The extent of shear penetration into the
obstruction, which is a primary determinant of residence time in the obstruction, scales
upon the drag length scale. Other relationships that link the strength of turbulence
and the ‘slip’ velocity at the top of the obstruction to the friction velocity (u∗ ) are also
evident. The relationships presented here provide predictive capability for flow and
transport in obstructed shear flows and suggest the possibility of a single framework
to describe such flows on all scales.
1. Introduction
Flows over permeable media (here termed ‘obstructed shear flows’) are ubiquitous in
the environment. Such permeable media include sediment beds, submerged vegetation
canopies and coral reefs in the aquatic environment and crop canopies, forests and
cities in the terrestrial environment. The physical, chemical and biological properties
of the fluid within the permeable medium often differ greatly from those of the fluid
outside it. It is therefore critical that we understand the rate at which momentum,
mass and heat are exchanged between the two regions. In the study of such varied
https://doi.org/10.1017/S0022112009992175
phenomena as pore water fluxes from sediments, the impact of forests on the global
carbon cycle or the amelioration of water quality by submerged vegetation, our
predictive capabilities are limited.
There is direct experimental evidence of Kelvin–Helmholtz-type vortices in flows
over vegetation canopies (figure 1). This instability, commonly observed in density-
stratified flows (Kundu & Cohen 2004) and regions where parallel streams merge
(a) (b)
Downloaded from https://www.cambridge.org/core. Montana State University Library, on 24 May 2021 at 01:55:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
(c) (d)
(Chu & Babarutsi 1988), is generated by an inflection point in the mean velocity
profile. In canopy flows, the vortices are responsible for the majority of vertical
transport across the top of the canopy (Gao, Shaw & Paw U 1989; Finnigan 2000;
Ghisalberti & Nepf 2009). The vortices engender strong similarity in the turbulent
properties of canopy flows, as demonstrated by Raupach, Finnigan & Brunet (1996).
For many other types of obstructed shear flow, the mean velocity profile is known to
contain an inflection point; the inflection point is located roughly at the top of the
obstruction (referred to hereafter as the ‘interface’). Inflectional profiles in flows over
https://doi.org/10.1017/S0022112009992175
a packed bed, a coral reef and an urban canopy are shown in figure 2. In the study of
many such systems, however, there is little evidence of the existence and predominance
of coherent interfacial vortices. A classical example is flow over sediment (i.e. packed)
beds. Although mixing at the sediment–water interface is almost invariably regarded
as occurring at molecular rates (Lorke et al. 2003), numerical simulations (figure 1d )
reveal that the flow near a highly permeable packed bed is dominated by a Kelvin–
Helmholtz-type instability (Breugem et al. 2006). The purpose of this paper is to
highlight a consistency of hydrodynamic/aerodynamic behaviour in obstructed shear
Obstructed shear flows 53
System Drag length scale, (CD a)−1 Height (h) required for CD ah > 0.1
Downloaded from https://www.cambridge.org/core. Montana State University Library, on 24 May 2021 at 01:55:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
3
Fluid 2.0
5
2
y (cm)
1.5
z/h
0
1 1.0
0 Interface –5
0.5
Porous media
–1
0 1 2 –10 0 10 0 0.2 0.4 0.6 0.8 1.0 1.2
u (cm s–1) velocity (cm s–1) uz/utop
Figure 2. Inflectional velocity profiles in flows over three permeable media with differing
length scales. (a) Oil flow over a packed bed of glass spheres (n = 0.41, (CD a)−1 ∼ O(1 cm),
Goharzadeh, Khalili & Jørgensen 2005, reproduced with the kind permission of the American
Institute of Physics) (b) Aquatic flow over a coral reef ((CD a)−1 ∼ O(10 cm), Reidenbach,
Koseff & Koehl 2009, reproduced with the kind permission of ASLO) (c) Atmospheric flow
over an urban canopy ((CD a)−1 ∼ O(10 m), Rotach et al. 2005, reproduced with the kind
permission of Springer). In all cases, the inflection point (the point of maximum velocity
gradient) and the interface (marked with a dot) are approximately coincident.
flows. In the absence of direct vortex visualization in many of these systems, the
similarity of their dynamics (i.e. the physical signature of the vortices) will be
demonstrated.
A permeable medium is characterized by its height (h) and its drag length scale,
(CD a)−1 . Here, CD is the drag coefficient of the medium and a is the frontal area
per unit volume, both averaged over the height of the medium. Typical drag length
scales are provided in table 1. The drag length scale is a more accurate measure of
the flow resistance within the obstruction than the porosity (n); the lower the drag
length scale, the greater the resistance. As shown in figure 3, the drag exerted by
the obstruction generates the unstable velocity profile. The ratio of the obstruction
height to the drag length scale (i.e. CD ah) must be greater than approximately 0.1
https://doi.org/10.1017/S0022112009992175
for obstructed shear flow behaviour to emerge (Nepf et al. 2007). If CD ah < 0.1,
the flow has the characteristics of a rough boundary layer, with no inflection point
in the mean velocity profile. The vortices that result from the unstable profile grow
downstream, reaching their equilibrium size a distance LT from the front of the
obstruction. This transition length scales upon the drag length scale (Ghisalberti &
Nepf 2009). Relative to the fully developed flow, turbulent mixing in the region x < LT
is diminished because of the reduced vortex scale but there can be strong vertical
advective fluxes due to the deceleration of fluid within the obstruction. Flexible media
54 M. Ghisalberti
(iii) Velocity profile develops
z U(z) (ii) As dU/dx < 0, inflection point, becoming unstable
there is an upward
Downloaded from https://www.cambridge.org/core. Montana State University Library, on 24 May 2021 at 01:55:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
δe
h
(i) Drag causes deceleration of (iv) Vortices reach final size a distance
fluid within obstruction Uh–U1 LT from front of obstruction
LT
Figure 3. The evolution of flow along a submerged permeable medium. The distance
required for full flow development LT scales upon the drag length scale of the obstruction.
relationship examines the structure of the turbulence, which is expected to differ from
that in a boundary layer.
2. Data extraction
Data were taken from flows over submerged aquatic vegetation canopies [in the
laboratory, rigid (Dunn, Lopez & Garcia 1996; Nepf & Vivoni 2000; Ghisalberti &
Nepf 2002; Poggi et al. 2004; Ghisalberti & Nepf 2004, 2006; Murphy, Ghisalberti
& Nepf 2007) and waving (Ghisalberti & Nepf 2002, 2006)], terrestrial vegetation
canopies [in the field (Shaw et al. 1974; Wilson et al. 1982; Baldocchi & Meyers 1988;
Amiro 1990a,b; Gardiner 1994; Katul & Albertson 1998; Kruijt et al. 2000; Novak
et al. 2000; Yi et al. 2005; Su et al. 2008) and laboratory (Seginer et al. 1976; Raupach,
Coppin & Legg 1986; Brunet, Finnigan & Raupach 1994; Novak et al. 2000)], urban
canopies [field (Oikawa & Meng 1995; Louka, Belcher & Harrison 2000; Rotach
et al. 2005) and laboratory (Macdonald 2000; Cheng & Castro 2002; Kastner-Klein
& Rotach 2004)], coral reefs [laboratory (Lowe, Koseff & Monismith 2005; Lowe
2005; Reidenbach, Koseff & Monismith 2007)], shallow flows horizontally adjacent
to (rather than over) vegetation stands [laboratory (Pasche & Rouvé 1985; White &
Nepf 2007)] and flows over DPM [laboratory (Gupte & Advani 1997; Goharzadeh
et al. 2005; Agelinchaab, Tachie & Ruth 2006)]. The term ‘dense porous medium’ is
used here to describe tightly packed obstructions ((CD a)−1 . 1 cm) and includes
packed granular beds, fibrous beds and rod arrays. In this study, these media
are treated as an analogue of sediment beds in aquatic systems. The rigid aquatic
vegetation data includes unpublished results from high-density arrays.
Some restrictions were placed upon the data used. Firstly, the obstruction had to
be sufficiently dense (or tall) to generate an inflection point in the velocity profile.
As discussed in § 1, this is true if CD ah > 0.1. For real systems, this is not an overly
restrictive condition (table 1). In the particular comparison of δe and (CD a)−1 , the
bed could not arrest vortex penetration. As the scaling coefficient in the relationship
δe ∼ (CD a)−1 was expected to be roughly 0.25 in the absence of bed effects (Nepf et al.
2007), only obstructions with CD ah > 0.25 were used in that comparison. Secondly,
aquatic flows where the free surface had a dynamic impact were not considered.
In shallow flows, the free surface can restrict vortex growth and strength (Nepf
& Vivoni 2000). To first order, roughly two-thirds of the vortices exist above the
canopy in vegetated flows (Ghisalberti & Nepf 2004). So, denoting the flow depth
as H , only flows for which CD a(H − h) > 0.5 were considered. Thirdly, flows with
waving vegetation canopies were used in the third and fourth comparisons only,
where definition of the obstruction height is not critical. Finally, flows with significant
https://doi.org/10.1017/S0022112009992175
density stratification were not considered. Only atmospheric flows classified as ‘near
neutral’ (for which h/|L| 1, where L is the Monin–Obukhov length) were included.
In all, data from 109 flows were used.
The parameters involved in the scaling of shear penetration into the obstruction
(namely CD , a and δe ) can be difficult to estimate from the literature. Here, I lay
out the method of estimating these three parameters when they were not explicitly
provided; this is particularly tricky for DPM.
56 M. Ghisalberti
2.1. Drag coefficient
Measurements of CD in an array of obstacles are surprisingly inconsistent. This is
Downloaded from https://www.cambridge.org/core. Montana State University Library, on 24 May 2021 at 01:55:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
it was taken as the distance over which the velocity gradient decays to 10 % of its
interfacial value.
2.4. Dense porous media
In the case of DPM, the combination of dense packing and low Reynolds numbers
make the estimation of CD virtually impossible. Furthermore, the shear penetration
into packed beds is typically of the order of the bed element diameter (Goharzadeh
et al. 2005). The drag length scale ((CD a)−1 ) is a bulk parameter, defined over many
Obstructed shear flows 57
(a) (b)
1
101 δe ≈ 3 (CD a)
–1
101 Uh – U1 ≈ 2.6u*
Downloaded from https://www.cambridge.org/core. Montana State University Library, on 24 May 2021 at 01:55:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
Uh – U1 (m s–1)
100 100
δe (m)
10–1 10–1
10–2 10–2
10–3 10–2 10–1 100 101 102 10–3 10–2 10–1 100
(CD a)–1 (m) u* (m s–1)
(c) (d) 1.6
wrms ≈ 1.1u* 1.4 wrms ≈ 0.6 urms
100
1.2
wrms (m s–1)
wrms/urms
1.0
10–1
0.8
0.6
10–2 0.4
0.2
0
10–3 10–2 10–1 100 10–1 100 101
u* (m s–1) CDah
Figure 4. Four key dynamic relationships in obstructed shear flows. The data are taken from
several systems: experimental flows over rigid model aquatic vegetation ( ), waving model
aquatic vegetation ( ), model terrestrial vegetation ( ), model urban canopies ( ), coral reefs
( ) and DPM ( ), experimental flows horizontally adjacent to vegetation stands ( ) and real
flows over terrestrial vegetation ( ) and urban canopies ( ). For each of the three relationships
presented in (a), (b) and (c), R 2 > 0.95 in log space. The dashed lines represent the 90 %
prediction interval.
inter-element spacings within the obstruction. Even for a perfectly packed bed, the
resistance imparted on the flow will vary significantly over the scale of a single bed
element. Thus, a true comparison of δe and (CD a)−1 is not possible for DPM. The
comparison that can be made is that between δe and a −1 , where a nominal value
of a is used for packed beds (namely, that in (2.2)). In effect, a drag coefficient of
unity has been assumed for the purpose of drawing comparison amongst DPM.
Importantly, however, the similarity of shear penetration behaviour between DPM
and other obstructed shear flows cannot be examined.
4. Distinguishing features
In the context of this study, the ideal obstructed shear flow is deep with a long,
uniform and comparatively sparse (n > 0.9) permeable medium. In such a flow, the
free surface does not impact vortex growth, the obstruction height and density are
easily defined and the Reynolds number of the interfacial flow is high (allowing
reasonable estimates of CD ). While figure 4 demonstrates similarities across the range
of obstructed shear flows, each system has peculiarities which distinguish it from the
ideal. Such distinguishing features are discussed below.
4.1. Dense porous media
In aquatic flows, sediment beds can be sufficiently dense so as to render the interfacial
flow laminar. Indeed, the Reynolds number (Re = Uh (CD a)−1 /v, where v is fluid
viscosity) of each of the nine DPM studies discussed here is less than 1. At such
low Reynolds numbers, it is possible that viscosity damps the vortex instability (in
contrast, Re ∼ O(100) in the numerical simulation in figure 1d ). Certainly, a quadratic
drag law, where CD is O(1) and independent of Re, will not strictly apply. Figure 4(a)
shows that the penetration of shear into DPM (often referred to as the Brinkman
layer thickness) is O(1/3a). This is also true in the other obstructed shear flows
studied here, for which CD ∼ O(1).
The idea of a diffusive boundary layer above sediment beds, where diffusion occurs
https://doi.org/10.1017/S0022112009992175
at molecular rates (e.g. Lorke et al. 2003), is common and can be practically applied
to predict solute fluxes from the sediment. This analysis does not prove the existence
of coherent vortices at the sediment–water interface. However, in combination with
the modelling work of Breugem et al. (2006) (figure 1d ), it does put forth the
possibility that the concept of a diffusive boundary layer might be flawed in some cases
(particularly for highly porous beds at high Reynolds number). That is, interfacial
mixing may in fact be governed by coherent structures rather than by Brownian
motion. Importantly, the velocity profiles of all DPM studied here (which have
Obstructed shear flows 59
porosities as low as 0.4) contain the necessary inflection point. Furthermore, linear
stability analysis suggests that flow over a permeable medium will be unstable (in
the absence of viscosity) for all non-zero porosities (White & Nepf 2007). However,
Downloaded from https://www.cambridge.org/core. Montana State University Library, on 24 May 2021 at 01:55:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
the lack of DPM data points in figure 4(b–d ), along with an inability to accurately
quantify CD for DPM, prevents further analysis. It is interesting to note that the
results of direct numerical simulation of flow over a highly porous packed bed
(Breugem et al. 2006, n = 0.95) agree very well with figure 4. Using peak values of
r.m.s. velocities, the coefficients in the relationships in figure 4(b–d ) are 2.7, 1.1 and
0.6 (respectively) in that flow. However, the impact of porosity on (a) the potential
for viscous damping of the instability and (b) the drag coefficient of a DPM is not
well understood. Indeed, the structure of interfacial turbulence in flows over DPM
requires a greater research focus.
4.2. Vegetation canopies
As mentioned, vegetation canopies (both aquatic and terrestrial) can have significant
flexibility and exhibit the dynamically important ho-/monami. While the waving
canopy data (brown triangles) follow the general trends in figure 4, it is difficult to
determine or predict key characteristics, such as CD or h, of a canopy whose geometry
oscillates over time. Also, due to light requirements, submerged aquatic vegetation is
typically found in shallow regions (Chambers & Kalff 1985; Duarte 1991), where the
free surface may restrict vortex development (Nepf & Vivoni 2000).
4.3. Urban canopies
Urban canopies are rarely uniform, with building height and spacing varying
significantly. Uniform patches of buildings may be much smaller than LT , which
would be O(100 m) for a typical urban canopy (Ghisalberti & Nepf 2009). This
would result in a flow with diminished vortex penetration, relative to the trend line in
figure 4(a). Indeed, the two urban canopy data points in this figure (black triangles)
fall below the general trend.
While each system has unique features, the common inflectional velocity profile
creates an inherent similarity among obstructed shear flows.
The author would like to thank Greg Ivey and Brett Branco for their insight and
feedback during the preparation of this manuscript.
REFERENCES
Agelinchaab, M., Tachie, M. F. & Ruth, D. W. 2006 Velocity measurement of flow through a
model three-dimensional porous medium. Phys. Fluids 18, 017105.
Amiro, B. D. 1990a Comparison of turbulence statistics within three boreal forest canopies. Bound.-
Layer Meteorol. 51, 99–121.
Amiro, B. D. 1990b Drag coefficients and turbulence spectra within three boreal forest canopies.
Bound.-Layer Meteorol. 52, 227–246.
Baldocchi, D. D. & Meyers, T. P. 1988 Turbulence structure in a deciduous forest. Bound.-Layer
https://doi.org/10.1017/S0022112009992175
University.
Lowe, R. J., Koseff, J. R. & Monismith, S. G. 2005 Oscillatory flow through submerged canopies:
1. Velocity structure. J. Geophys. Res. 110.
Macdonald, R. W. 2000 Modelling the mean velocity profile in the urban canopy layer. Bound.-
Layer Meteorol. 97, 25–45.
Murphy, E., Ghisalberti, M. & Nepf, H. 2007 Model and laboratory study of dispersion in flows
with submerged vegetation. Water Resour. Res. 43, W05438.
Nepf, H., Ghisalberti, M., White, B. & Murphy, E. 2007 Retention time and dispersion associated
with submerged aquatic canopies. Water Resour. Res. 43, W04422.
Obstructed shear flows 61
Nepf, H. M. & Vivoni, E. R. 2000 Flow structure in depth-limited, vegetated flow. J. Geophys. Res.
105 (C12), 28547–28557.
Nezu, I. & Nakagawa, H. 1993 Turbulence in Open-Channel Flows. Balkema.
Downloaded from https://www.cambridge.org/core. Montana State University Library, on 24 May 2021 at 01:55:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
Novak, M. D., Warland, J. S., Orchansky, A. L., Ketler, R. & Green, S. 2000 Wind tunnel and
field measurements of turbulent flow in forests. Part I: uniformly thinned stands. Bound.-Layer
Meteorol. 95, 457–495.
Oikawa, S. & Meng, Y. 1995 Turbulence characteristics and organized motion in a suburban
roughness sublayer. Bound.-Layer Meteorol. 74, 289–312.
Pasche, E. & Rouvé, G. 1985 Overbank flow with vegetatively roughened flood plains. J. Hydraul.
Engng – ASCE 111 (9), 1262–1278.
Poggi, D., Porporato, A., Ridolfi, L., Albertson, J. D. & Katul, G. G. 2004 The effect of
vegetation density on canopy sublayer turbulence. Bound.-Layer Meteorol. 111, 565–587.
Raupach, M. R., Coppin, P. A. & Legg, B. J. 1986 Experiments on scalar dispersion within a model
plant canopy. Part I: the turbulence structure. Bound.-Layer Meteorol. 35, 21–52.
Raupach, M. R., Finnigan, J. J. & Brunet, Y. 1996 Coherent eddies and turbulence in vegetation
canopies: the mixing-layer analogy. Bound.-Layer Meteorol. 78, 351–382.
Reidenbach, M. A., Koseff, J. R. & Koehl, M. A. R. 2009 Hydrodynamic forces on larvae
affect their settlement on coral reefs in turbulent, wave-driven flow. Limnol. Oceanogr. 54 (1),
318–330.
Reidenbach, M. A., Koseff, J. R. & Monismith, S. G. 2007 Laboratory experiments of fine-scale
mixing and mass transport within a coral canopy. Phys. Fluids 19, 075107.
Rotach, M. W., Vogt, R., Bernhofer, C., Batchvarova, E., Christen, A., Clappier, A., Feddersen,
B., Gryning, S.-E., Martucci, G., Mayer, H., Mitev, V., Oke, T. R., Parlow, E., Richner,
H., Roth, M., Roulet, Y.-A., Ruffieux, D., Salmond, J. A., Schatzmann, M. & Voogt, J. A.
2005 BUBBLE – an urban boundary layer meteorology project. Theor. Appl. Climatol. 81,
231–261.
Seginer, I., Mulhearn, P. J., Bradley, E. F. & Finnigan, J. J. 1976 Turbulent flow in a model
plant canopy. Bound.-Layer Meteorol. 10, 423–453.
Shaw, R. H., den Hartog, G., King, K. M. & Thurtell, G. W. 1974 Measurements of mean wind
flow and three-dimensional turbulence intensity with a mature corn canopy. Agric. Meteorol.
13, 419–425.
Su, H.-B., Schmid, H. P., Vogel, C. S. & Curtis, P. S. 2008 Effects of canopy morphology and
thermal stability on mean flow and turbulence statistics observed inside a mixed hardwood
forest. Agric. Forest Meteorol. 148, 862–882.
White, B. L. & Nepf, H. M. 2007 Shear instability and coherent structures in shallow flow adjacent
to a porous layer. J. Fluid Mech. 593, 1–32.
Wilson, J. D., Ward, D. P., Thurtell, G. W. & Kidd, G. E. 1982 Statistics of atmospheric
turbulence within and above a corn canopy. Bound.-Layer Meteorol. 24, 495–519.
Yi, C., Monson, R. K., Zhai, Z., Anderson, D. E., Lamb, B., Allwine, G., Turnipseed, A. A. &
Burns, S. P. 2005 Modeling and measuring the nocturnal drainage flow in a high-elevation,
subalpine forest with complex terrain. J. Geophys. Res. 110, D22303.
https://doi.org/10.1017/S0022112009992175